Myanmar Sub-national Budget Brief · Ildrim Valley is an economist on the Overseas Development...

Myanmar Sub-national Budget Brief

2

About the authors

Ildrim Valley is an economist on the Overseas Development Institute (ODI) Fellowship Scheme. He has previously

worked as a research consultant at the World Bank’s Development Research Group. Hein Aung Kyaw is a program

coordinator with Renaissance Institute. He has previously worked for the Central Bank of Myanmar.

Focus and approach

The Renaissance Institute (RI) developed this report as part of its efforts to support Myanmar’s public finance reforms.

The brief is based on nearly two years of intensive and ongoing engagement with budget departments across Myanmar.

It contributes to RI’s budget brief series, which aims to build a better understanding of sub-national finance, by

summarizing findings from direct interactions with government officials and collected data throughout RI’s support

to sub-national governments.

Acknowledgment

The authors would like to thank the budget departments and the respective state and region government authorities

in Bago, Kayin, Ayeyarwaddy and Nay Pyi Taw for their assistance and support in undertaking this analysis. They are

not responsible for any of the conclusions in this report. Any errors are the responsibility of the authors. Thanks also

to colleagues in the Renaissance Institute (in particular Chaw Su Min Han, Andrew Wilson, and Khin Moe Myint) and

the Asia Foundation (in particular Richard Batcheler). We appreciate the support and guidance of U Myo Myint and

U Soe Win of Renaissance Institute. This research was funded by the governments of the United Kingdom, Australia,

and Switzerland with support from The Asia Foundation. The views expressed in this publication are the author's

alone and are not necessarily the views of the funding partners.


3

Background: Myanmar has taken steps to transition from a centralized to a decentralized system of government

based on the 2008 Constitution, which established 14 sub-national governments (seven states and seven regions).

Sub-national institutions have some devolved political authority with a range of finance and administrative functions.

The new sub-national governments (SNGs) and sub-national institutions are playing an increasingly prominent role

in spending and managing of public finance, the delivery of services and ultimately economic growth.

A path to better governance of these new institutions will have to address their public financial management (PFM).

Systems underpinning the management of public funds are crucial to how decisions are made about essential services

and infrastructure that citizens use in their daily life: which roads are built, or how waste is collected. These are all

fiscal policy choices that require institutions to support their implementation. PFM revolves around how fiscal

institutions work and will be essential to emerging sub-national institutions in Myanmar’s states and regions.

Objective: This report aims to provide an overview of Myanmar’s sub-national public finance management. It

focuses on sub-national budgets as the most important local policy document. It also touches on broader PFM

institutions, to suggest areas for further examination and highlight the significance of cash management, accounting

practices, fiscal risks and intergovernmental relations. PFM is also critical for oversight mechanisms and transparency.

Good PFM contributes to building trust, makes the government ultimately accountable for its actions and choices.

Citizens, firms, and investors all want to know whether public money is used effectively and in line with the goals and

priorities that the society has set. Modern PFM is very much related to fiscal reporting requirements, transparency

standards, and the availability of information available to the public in an accessible and meaningful way. The main

outcome is that it contributes to building trust between the government and other economic players, to create a

favorable setting for productive, efficient, and inclusive economic activities. This reflects the spirit of this document

to shed light on issues related to management of public money, how it’s used and collected.

Summary: The sub-national governments in Myanmar are playing an increasingly larger role in general government

finance. Their share of public sector spending has increased from around 6 percent in 2012-13 to almost 12 percent

in 2017-18. A relatively large share of this spending is used for capital investments.

Intergovernmental fiscal transfers play a crucial role in state/region finance, not least because they finance the majority

of local budgets. Recent changes to a more rules-based system have also had impact on transparency of local financing,

macro-sustainability and clearer expectations of available resources. Going forward, they may yet offer a deeper scope

as a policy tool for the Union Government.

Collection of own-source revenue is concentrated in Yangon and Mandalay, where they finance nearly half of their

budgets, a contrast to around only 10 percent of budgets in 12 other states/regions. Eight sub-national taxes play a

relatively small part in local budgets with collection fragmented across 4 different ministries. On the other hand, non-

tax revenue accounts for the majority of own-source revenue. The exact nature of non-tax receipts is blurred by

transparency standards and budget accounting practices.

Sub-national governments have made progress in improving public access to key fiscal information. However, there

is room for more significant progress, in both catching up with the Union Government practices and introducing

other transparency initiatives. The quality of public information is weakened by inconsistent accounting classification,

particularly around capital investments. This in turn limits control over the quality of expenditure and general fiscal

oversight.


4

Table of contents

Acronyms 5

Background 6

A. Aggregate picture and fiscal developments 7

B. Revenue arrangements and performance 9

C. Expenditure 21

D. Institutions and management 26

E. Fiscal risks 31

F. Concluding remarks 32

References 33

Appendix 34


5

Acronyms

BE Budget Estimate

BOT Build-Operate-Transfer

CDF Constituency Development Fund

DOH Department of Highways

DRRD Department of Rural Roads Development

FA Fund Account

FY Fiscal Year

GAD General Administration Department

GDP Gross Domestic Product

GRF General Reserve Fund

IMF International Monetary Fund

IRD Internal Revenue Department

JICA Japan International Cooperation Agency

LTO Large Taxpayers Office

MCDC Mandalay City Development Committee

MEB Myanma Economic Bank

MOC Ministry of Construction

MOPF Ministry of Planning and Finance

MTFF Medium-Term Fiscal Framework

ODA Official Development Assistance

OECD Organization for Economic Cooperation and Development

OMI Open Myanmar Initiative

PEFA Public Expenditure and Financial Accountability

PFM Public Financial Management

PPP Public-private partnership

SAA Self-administered Areas

SEE State Economic Enterprises

SOE State Owned Enterprises

SNG Sub-national government

YCDC Yangon City Development Committee

UCLG United Cities and Local Governments

USD United States Dollar


6

Background

1. Myanmar has taken steps to transition from a centralized to a more decentralized system of

government based on the 2008 Constitution. In a departure from a highly centralized structure, 14 sub-national

governments (seven states and seven regions) were established, with partially elected parliaments (Figure 1).1 In

addition to their devolved political authority, a range of finance and administrative functions were ceded to this newly

minted level of government. Overall, the sub-national governments (SNGs) are playing an increasingly prominent

role in the management and spending of public finance and the delivery of services.

2. This report focuses on public financial management of the 14 sub-national governments. 2 It

summarizes composition of sub-national budgets, namely: what local receipts help fund sub-national budgets; on

what basis are central fiscal transfers utilized to close fiscal gaps; overview of what administrative units manage said

funds. It also provides a brief overview of the management of public funds, transparency, cash management, and

fiscal risks. Although districts and townships play an important role in influencing the quality of public financial

management throughout the budget cycle, their detailed examination is beyond the scope of this report.

Figure 1: Government hierarchy

3. Administrative units in sub-national budgets: sub-national spending and receipts are managed by

departments within the Union line ministries. The sole exception to this are Municipal level offices, Development

Affairs Organizations (DAOs) and City Development Committees in Yangon and Mandalay (YCDC and MCDC),

which do not have a Union level ministry and occupy a unique position in an otherwise highly centralized state.

Although not recognized as a formal level of government, municipal offices play an important role in the generation

of sub-national revenue and service provision (for the summary of administrative units on sub-national budgets, see

appendix). Budget department offices in each state and region play a key role, providing horizontal and vertical

coordination during the preparation stage of the budget cycle, as well as a degree of oversight and reporting

throughout the budget cycle.

1 States and regions form a second-tier government and are constitutionally equivalent. Regions have been historically populated by an ethnic Bamar

majority. States are recognized for their ethnic minority dominated populations. Seven states are Chin, Kachin, Kayah, Kayin, Mon, Rakhine and Shan. Seven regions are Ayeyarwaddy, Bago, Magwe, Mandalay, Sagaing, Tanintharyi and Yangon. 2 Self-administered areas (SAA) and Union territory of Nay Pyi Taw are not explored in the report.

Union Government

14 States/Regions

67 Districts

330 Townships


7

A. Aggregate picture and fiscal developments

4. State and region government budgeted expenditure has grown significantly since the 2012-13 FY.

Expenditure has nearly tripled since first enacted budgets in 2012-13, increasing from 864,122 million Kyat to

2,474,942 million Kyat between 2012-13 and 2017-18 fiscal years. As a proportion of general government spending,

sub-national spending increased from around 6 percent in 2012-13 to nearly 12 percent in 2017-18 (Figure 2).

5. Sub-national spending remains modest at about 50,000 Kyat per person on average across 14 states/regions.

In relative terms, the state/regions’ share of general government expenditure is below that of other Asian economies,

above Thailand’s 10 percent but below countries like Vietnam (45 percent), Indonesia (35 percent) and Bangladesh

(15 percent) (Figure 3). Expenditure composition is explored in section C.

6. Sub-national governments are more reliant on central transfers relative to international comparators.

About 80 percent of sub-national budgets in Myanmar are financed by the central Union Government – with variation

across states and regions (Figure 4A). 3 With the exception of a few countries and large urban centers, local

governments commonly face fiscal shortfalls and intergovernmental transfers to SNGs are a fact of life in public

finance. There isn’t an ideal or preferred composition of SNG fiscal resources. Although, according to OECD, across

a sample of 90 countries, the share of resources financed by grants and subsidies tend to be lower – and local tax

collection higher – for higher income countries relative to low-income countries (Figure 4B). Myanmar fits at the far

3 As high as 97 and 99 percent in Tanintharyi region and Chin state, respectively. On the opposite side of the spectrum, around 50 percent of budgets in

Yangon and Mandalay financed by local sources of revenue.

0

500,000

1,000,000

1,500,000

2,000,000

2,500,000

3,000,000

0

5,000,000

10,000,000

15,000,000

20,000,000

25,000,000

30,000,000

2012-13 2013-14 2014-15 2015-16 2016-17 2017-18 2012-13 2013-14 2014-15 2015-16 2016-17 2017-18

Figure 2: Expenditure across levels of government2012-2018 in million Kyat

State/Region expenditure Union expenditure

Source: Union Citizen's Budget, MOPF; RI staff estimates

6.4%8.3%

11.8%

State/Region expenditure as % of general government spending

Figure 3: Sub-national spending as share of

general government expenditure International comparison

Source: Adapted from Minoletti, 2016; International figures (2009) taken from Dickenson-Jones, De and Smurra, State and Region Public

Finances in Myanmar, p.44; Myanmar figure (2017/18) from Union Citizen’s Budget and MOPF data.

60%

45% 45%

35% 33%

25%

15% 12% 10%

Japan Vietnam Korea Indonesia Pakistan Philippines Bangladesh Myanmar Thailand


8

end, with a higher share of resources financed by grants than the average amongst the low-income countries. The

following section (B) provides more detail on the composition of SNG revenue in Myanmar.

Figure 4: Breakdown of sub-national governments revenue by category, % of SNG revenue

7. Medium Term Fiscal Framework (MTFF) reforms: MTFF reforms in Myanmar aim to foster fiscal

discipline and improve strategic resource allocation by providing expenditure ceilings to government agencies. As

part of these reforms, there have been significant changes to sub-national financing – more specifically, to

intergovernmental fiscal transfers. Traditionally, Union transfers to state/region governments have been negotiated,

functioning as “deficit grants”, often adjusted to help achieve balanced budgets. As part of the medium-term fiscal

framework (MTFF) reforms, since 2015-16 FY a large share of fiscal transfers has been rules-based. These changes,

and the fiscal transfers more broadly, are discussed in more detail in the revenue section (section B).

8. Budget calendar: One of the most significant recent developments is a shift in Myanmar's budget calendar,

from April-March to October-September. To accompany the change a 6-month interim budget was introduced to

cover the period bridging the end of the 2017-18 FY on March 31, 2018, and the new 2018-19 fiscal year, to

commence on October 1, 2018. The difficulties in implementing infrastructure projects hampered by the onset of the

rainy season (during the months of June - October) is the commonly cited rationale behind the change. The interim

budgets are not included in this document, although changes to the calendar are broadly discussed.4

9. Institutions: Sub-national institutions that relate to PFM (e.g. the budget departments, Hluttaws5), have been

formed less than a decade ago. To those outside the government, there are still important question marks around

accounting practices, flow of funds and information both horizontally and vertically. This report highlights these areas

as important areas of knowledge for the purposes of external technical support to the government. A deeper inquiry

of these will allow government partners and donors to foster a better understanding of Myanmar’s new institutions

and provide meaningful support to improving local governance, integration and coordination of activities across levels

of government.

4 Some coverage of the interim budgets for individual states/regions can be found in RI budget briefs - https://rimyanmar.org/en/publication-

category/budget-briefs 5 Hluttaws are Myanmar’s (partially) elected parliaments. Each state and region has its own Hluttaw.

69%

9%3%

19%

0%

20%

40%

60%

80%

100%

Myanmar

Other revenue SNG tax

Tax revenue sharing General grant transfer

Source: MoPF; RI staff estimates

4A: 2016/17 FY, across 14 SNGs 4B: 2013, by income level

Source: OECD/UCLG, 2016


9

B. Revenue arrangements and performance

10. Background: Revenue in Myanmar’s SNGs is classified under 3 categories: current, capital and financial. The

majority of all revenue across sub-national governments is budgeted as current, which is the focus of this section.

Capital and financial revenue are almost entirely concentrated in Yangon and Mandalay and are not explored in much

detail in this brief.

11. Current revenue can be partitioned into i) own sourced revenue collection (categorized as tax and non-tax)

and ii) Union transfers (Table 1).6 What this report refers to as "own" revenue excludes the Union general grant

transfer, Community Development Fund (CDF), Union collected shared taxes, and other ad-hoc transfers prior to

2016-17 FY. Although it’s not uncommon to characterize centrally shared taxes as sub-national own revenue, this

report treats them as transfers. The alternative characterization would less accurately reflect the real level of tax

autonomy of SNGs. Regions and states don't have a lot of leeway over rates and bases for shared taxes.

Table 1: Current revenue breakdown for SNGs in Myanmar

Fiscal transfers from Union Own revenue

Non-tax Tax

Union general-

purpose grant

transfer

Constituency

Development Fund

(CDF), other ad-hoc

transfers before

2015-16 FY

Union tax revenue sharing:

- commercial tax

- income tax

- special goods tax

- stamp duty tax

Fees,

penalties,

monopoly

licenses etc.

Wheel tax, property tax, excise tax,

mineral, land tax, water and

embankment tax, tax on the

extraction of forest products, tax

on fisheries Note: Tax revenue refers to eight items labeled as taxes on sub-national budgets

12. SNGs outside of Yangon and Mandalay are firmly reliant on Union transfers (Figure 5A). About half

of the revenue in Yangon and Mandalay comes from transfers. The figure is close to 90 percent for the rest of the

country. In 2016-17 FY, Yangon and Mandalay have derived some of their revenue from the sale of capital assets –

14 and 16 percent of their fiscal resources, respectively. While it makes a significant share of their income, capital

income stems from the one-off sale of assets and hence does not provide a sustainable source of revenue (Figure 5B).

6 Available budget formats for some individual states/regions pose a challenge in separating own source non-tax revenue from some of the Union transfers –

with items both bundled together under “other current revenue” – which consequently blurs the picture of SNG finances.

Figure 5: Breakdown of SNG fiscal resources By type, 2016-17 BE


Notes: SNG = Sub-national Government; CDF = Constituency Development Fund; YGN = Yangon; MDL = Mandalay; BE = Budget Estimate (enacted by legislature)

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

All YGN &MDL

OtherSNGs

5A - Share of total resources

0

100,000

200,000

300,000

400,000

500,000

600,000

Kac

hin

Kay

ah

Kay

in

Ch

in

Saga

ing

Than

inta

ryi

Bag

o

Mag

gway

Man

dal

ay

Mo

n

Rak

khin

e

Yan

gon

Shan

Aye

yarw

add

y

5B - Across 14 states/regions, million Kyat

Financial revenue

Capital revenue

Other non-tax current revenue

SNG Tax

CDF

Tax-sharing revenue

General grant transfer

Transfers from Union

Own currentrevenue

Mag

we

Rak

hin

e


10

13. Available fiscal resources per citizen vary across states and regions. Figure 6 illustrates available budgeted

fiscal resources for each state and region for 2017-18 FY. They include both own-sourced revenues and fiscal transfers

from Union Government. Chin state has the highest per capita resources, with about 297,000 Kyat per person. On

the opposite side of the scale is Ayeyarwaddy region with under 24,000 Kyat per person. These variations are largely

an outcome of how intergovernmental fiscal transfers are administered, which are explored in section B2. The

composition of expenditure is explored in section C.

14. The following section reviews the level, composition, and arrangements of revenue: with a focus on both own-

sourced current revenues and fiscal transfers from the Union Government.

B.1 Own revenue

15. Background: State/Region governments collect both tax and non-tax revenues from fees, licenses, and more,

within the powers under schedule 5 of the 2008 Constitution. There is scope for deeper inquiry into local sources of

revenue. Specifically, into the legal framework, policy and administration elements of local collection – some of which

are shaped by pre-Constitution laws. Moreover, schedule 5 items are not easily identified in local budgets. It’s possible

that not all of the schedule 5 items are earmarked for sub-national budgets and are still budgeted at the Union level

(see appendix for the list of schedule 5 items and links to SNG budgets).

16. Amendments to the constitution in 2015 added a list of taxes to schedule 5, to be potentially collected by

states and regions. While it had no immediate impact on SNG finance, the amendment acts as a placeholder for

subsequent Union laws that could lead to considerable changes. For example, the list includes taxes on natural

resources and customs; decentralization of which may have significant implications for horizontal equity between

locations.7

17. Tax vs. non-tax revenue. In practice, the distinction between tax and non-tax sub-national revenue is blurry

and available budget formats pose challenges to meaningful analysis. This report identifies 8 tax items (Table 1) –

those regularly labeled as such under sub-national budgets. It could be argued that some of these taxes function more

as license fees for running particular business operations (e.g. tax on fisheries, excise tax). Non-tax revenue makes the

majority of own-source revenue (Figure 7A). Available budget formats conceal their detailed breakdown, but more

broadly, they include: various fees from renting of government assets, fines, seizures, selling of business licenses and

more.

7 Section 3 of the Law Amending the Constitution of Republic of the Union of Myanmar, The Pyidaungsu Hluttaw Law No. 45, 22 July 2015.

0

50,000

100,000

150,000

200,000

250,000

300,000

Figure 6: Available fiscal resources Kyat per capita by states/regions, 2017-18 BE

Source: MOPF, 2014 census; RI staff estimatesNote: BE = Budget Estimate (approved by legislature)


11

18. Indirect taxation: sub-national collection presents a distortionary mix of revenues. Government levies

everywhere influence decision-making to some extent, and tend to introduce some distortions. A well-defined tax

policy would aim to balance the need to fund government while keeping negative market distortions to a minimum.

In Myanmar’s SNGs, both tax and non-tax sources of revenue provide multiple examples of distortions that impose

arbitrary and unequal burdens of similar members of society. Bissinger (2016) suggests that this contributes to a

regressive revenue structure, with a study focus on non-tax sources like the municipal auction licenses. Taxes too, like

those on fisheries, may function like business license fees and can contribute to regressive outcomes. The requirement

is often to pay a fee/tax in advance to engage in an economic activity. These are generally paid by businesses, though

the cost can often be passed on to consumers. Such system disadvantages credit-poor small businesses and may

disproportionally affect women (Chan et al., 2018).

19. Informal taxation: When talking about sub-national taxes and tax reform in Myanmar it is important to look

beyond formal budget classifications and levies collected by the government. Decades of military rule have built a

degree of distrust between businesses / citizens and the state (Bissinger, 2016). This affects what taxes are collected

and how the formal collection is administered.8 In addition, McCarthy (2016) suggests that lack of state-supported

programs have helped develop localized informal mechanisms or risk-sharing, and public goods provision across

Myanmar. He concludes that non-state welfare systems are essential to social protection, shaping expectations of the

state and placing a significant burden on households to contribute to these institutions. Little is known about the

relative importance of these non-state networks and the extent of fiscal burdens on households when compared to

what government levies in tax and provides in support. This may have implications on tax reform that aims to reach

the broader Myanmar population. It also highlights the importance of a role that sub-national governments / offices

can play in developing location-specific strategies, building trust and understanding about the reforms before they

take place.

20. Almost all of own budgeted revenue – outside of the three Union transfers – is generated in Yangon

and Mandalay regions (about 85 percent in 2016-17 BE). In 2016-17, Yangon and Mandalay accounted for nearly

all of capital and financial revenue, 65 percent of tax revenue and about 80 percent of non-tax revenue (Figure 7B).

This highlights a wide gap in revenue performance between the 12 SNGs relative to Yangon and Mandalay.

21. Municipal offices collect the majority of SNG own revenue – 77 percent (Figure 8A). Most of this

revenue is collected by Yangon and Mandalay City Development Committees (YCDC and MCDC) from non-tax

sources like sale of capital assets, sale of monopoly auction licenses, various fees, and some financial revenue (see Box

8 For more detail see, Bissinger (2016)

Figure 7: Breakdown of own source fiscal resources By type, as % of state/region budgets, excluding Union transfers, 2016-17 BE

Source: MoPF; RI staff estimates Note: BE = Budget Estimate (enacted by legislature); YGN = Yangon; MDL = Mandalay; SNG = Sub-national Government; “Others” includes 12 states/regions outside of Yangon and Mandalay regions

22%

9%

16%

53%

7A - all 14 states/regions

Capital

Financial

SNG Tax

Non-tax: othercurrent revenue

0%

20%

40%

60%

80%

100%

Capital Financial SNG Tax Non-tax:other

currentrevenue

YGN & MDL Others

7B - Yangon and Mandalay vs. other SNGs


12

1).9 Sub-national government bodies (mostly SNG cabinet offices)10 account for about 17 percent of local revenue

through collection of various fees and penalties. Available budget formats limit analysis of non-tax revenue sources

for both municipal and SNG cabinet offices.

22. Overview of local tax arrangements in Myanmar: Tax assignments refer to items identified as local taxes

on sub-national budgets and with basis in schedule 5 of the constitution. However, formal revenue assignment may

not translate to full local autonomy. With greater tax autonomy state/region authorities would have the ability to

determine the rates (and potentially also the bases) for the constitutionally assigned taxes. Among locally assigned

taxes, SNGs have full autonomy over property tax (Lachlan and Hein, 2017). This is broadly in line with sound

assignment principles, with good practice to assign taxes with immobile bases (e.g. land and property) to local

authorities because there is no scope to shift location to avoid taxation.

23. Tax autonomy: There’s been lack of clarity in how local governments’ exercise discretion over tax policy, with

respect to tax base and rates over other schedule 5 taxes. However, in recent years several states/regions have passed

9 For more detail on municipal revenues and auction licenses see, “Local Economic Governance in Myanmar”, Jared Bissinger 2016 -

http://asiafoundation.org/publication/local-economic-governance-in-myanmar/ 10 Sub-national government Bodies are comprised of (1) SNG Cabinet Office (2) SNG Parliament (3) Court (4) Attorney General Office and

(5) Auditor Office.

Figure 8: SNG own-source revenue million Kyat

Source: Open Myanmar Initiative; RI staff estimates Notes: Figure 8A: Municipal administrative units includes Development Affairs Organizations, Yangon and Mandalay City Development Committees (CDCs); Sub-national government Bodies are comprised of (1) SNG Cabinet Office (2) SNG Parliament (3) Court (4) Attorney General Office and (5) Auditor Office; Figure 8B includes data for 11 states/regions; excludes Chin state, Rakhine state, Ayeyarwaddy region.

0

100,000

200,000

300,000

400,000

500,000

600,000

2016-17

8A - by administrative units

Others

Sub-nationalgovernment bodies

Ministry of Home Affairs

Municipal - DAOs / CDCs

0

50,000

100,000

150,000

200,000

250,000

300,000

350,000

400,000

2013-14 2016-17

8B - by type

Loan income (YCDC)

Borrowed (MCDC)

Capital

Tax

Other income

Box 1: Sub-national borrowing

Sub-national borrowing is specified under the 2016 Public Debt Management Law. It indicates that states and regions can

borrow from both Myanmar and international lenders, subject to approval from the Union Cabinet and the Union Hluttaw

(Public Debt Management Law, Chapter 6, Article 20). To date, it appears no significant external borrowing has been

undertaken by Myanmar SNGs outside of Yangon and Mandalay.

Financial revenue for sub-national governments is exclusively budgeted under the municipal offices in Yangon and

Mandalay (YCDC and MCDC). Publicly available budget formats limit a more detailed inquiry, with aggregate financial

revenue figures budgeted as “loan income” and “borrowed”. It’s not entirely clear if loan income refers to a debt obligation

and/or income made from issuing loans. Conversations with Yangon MPs suggests that financial revenue relates to the

ODA loan issued by JICA, which covers the Greater Yangon Water Supply Improvement Project – handled by the YCDC.

For more detail on the JICA loan, see, https://www.jica.go.jp/english/news/press/2016/170301_01.html Public Debt Management Law: http://www.myanmar-law-library.org/law-library/laws-and-regulations/laws/myanmar-laws-1988-until-now/union-solidarity-and-development-party-laws-2012-2016/myanmar-laws-2016/pyidaungsu-hluttaw-law-no-2-2016-public-debt-management-law-burmese.html

buda

Financial revenue is budgeted as “loan income” and “borrowed” (Figure 8B).


13

their own tax laws altering the tax base and rates. According to media sources, Yangon has passed its own tax law

which is said to cover tax rates on property, land, fisheries and wheel tax.11 Kayin state passed the State Land Tax

Law raising tax rates on land from under 5 Kyat per acre to 300-700 Kyat per acre depending on type of land. The

law also clarifies tax base, timeframe, and administration responsible for collection. 12 These examples provide

suggestive evidence that SNGs can and do set their own rates. However, further examination is required for a wider

picture of autonomy over local tax policy. Currently, state and region governments have no autonomy over Union

collected shared revenues, which are thus treated in this report as fiscal transfers. Surcharges on national taxes are not

explored in Myanmar.

Table 2: Taxonomy of sub-national revenue arrangements

Local autonomy

Tax base Tax rate Administration

Revenue source Typical Myanmar Typical Myanmar Typical Myanmar

Own-revenue assignments (Schedule 5)

Property tax* and wheel tax ✓ ✓ ✓ ✓ Possibly ✓

Other Schedule 5 taxes ✓ Possibly ✓ Possibly Possibly ✗

Revenue sharing ✗ ✗ ✗ ✗ Possibly ✗

Surcharges on national taxes ✗ N/A ✓ N/A Possibly N/A

Source: Adapted from Fedelino and Ter-Minassian (2010) and World Bank (2015); *McDonald and Hein (2017); Note: N/A = Not Applicable

24. Sub-national tax performance since 2012-13 has been weak and growth relatively flat outside of Yangon

and Mandalay (Figure 9). Average collection of 8 taxes identified on SNG budgets in 14 states and regions has

increased from 1,000 Kyat in 2012-13 FY to 2,216 Kyat per person in 2017-18 FY, or about $1.5. There are however

significant variations – with collection as low as 458 Kyat ($0.3) in Rakhine and as high as 7,100 Kyat ($5) in Yangon.13

Annual growth in real terms averaged 13 percent in Yangon and Mandalay, and only about 3 percent across the other

12 states/regions.14

25. Tax to GDP: Weak tax performance outside of Yangon and Mandalay is reflected in the low share of taxes

relative to state/region economic output. Between 2013-14 and 2017-18 fiscal years, tax performance outside of

Yangon and Mandalay has not kept pace with GDP (Figure 10). During the same period, the ratio of tax to GDP has

increased only in Yangon and Mandalay by 0.1 and 0.4 of GDP, respectively.

11 https://www.imyanmarhouse.com/news/read/721694 12 Kayin State Land Tax Law – enacted in Kayin State Hluttaw Law No. 3/2017 13 Exchange rate used for USD conversation as of July 2018: 1 USD = 1416 Kyat. 14 Real terms - deflated using the IMF average consumer price index for Myanmar.

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

0

200

400

600

800

1,000

1,200

1,400

1,600

2012-13 2017-18 2012-13 2017-18

Figure 9: SNG per capita tax collectionKyat per person in 14 states/region, 2012-2018

Kachin Kayin Chin Sagaing Thanintaryi

Kayah Bago Maggway Mon Rakkhine

Shan Ayeyarwaddy Mandalay Yangon


Yangon

Mandalay

TanintharyiAyeyarwaddy

ShanRakhine

RakhineMagwe

https://www.imyanmarhouse.com/news/read/721694


14

26. Sub-national tax collection is fragmented across multiple administrative units (Figure 11A). Eight taxes

identified on sub-national budgets are collected by 4 different ministries. Property and wheel taxes are collected by

the municipal offices. General Administration Department (GAD.) under the Ministry of Home Affairs, collects 4

other taxes: land tax, excise tax, water and embankment tax, and mineral tax. Tax on extraction of forest products and

tax on fisheries are collected by Departments of Forestry (under the Ministry of Natural Resources and Environmental

Conservation) and Fisheries (under the Ministry of Agriculture, Livestock and Irrigation), respectively. Such

fragmented administration may pose challenges to tax reform, as policy and building of technical capacity would

require wider effort and coordination. Average tax collection of 2,216 Kyat per person in 2017-18 FY, or about $1.5,

appear even smaller when one considers that it’s collected by 4 different ministerial administrations. This raises a

question of whether such fragmentated of tax collection is worth the benefit (especially for taxes that may cause more

market distortion relative to what they bring in funds, paragraph 18).

27. Wheel, property and excise taxes are largest tax items for states/regions – 32, 28 and 24 percent of the

total, respectively. Although there are variations across states/regions: Fisheries for Ayeyarwaddy and Forest

production for Tanintharyi are the largest tax items, respectively (Figure 11A). Administratively, municipal offices bring

in 76 percent of all tax collection for Yangon and Mandalay; as compared to 29 percent in other SNGs (Figure 11B).

Outside of Yangon and Mandalay, GAD under the Ministry of Home Affairs collects the most tax with 37 percent of

all collection – majority via excise tax (for more on excise tax, see box 2).

Figure 11: Breakdown of SNG tax collection By item, 2016-17 BE

Source: MoPF; RI staff estimates; Notes: BE = Budget Estimate (enacted by legislature); Municipal offices include Development Affairs Organizations (DAOs), Yangon and Mandalay City Development Committee (YCDC, MCDC); GAD = General Administration Department (Ministry of Home Affairs). Tax on extraction of forest products and tax on fisheries are collected by departments of forestry and fisheries, respectively.

0%

2%

4%

6%

8%

Figure 10: SNG tax to GDP8 SNG taxes as % of state/region GDP

2013-14 2017-18


0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

5,000

Shan

Rak

khin

e

Mag

gway

Kay

in

Ch

in

Saga

ing

Kay

ah

Kac

hin

Bag

o

Mo

n

Aye

yarw

add

y

Than

inta

ryi

Man

dal

ay

Yan

gon

All

11A - Kyat per capita

Wheel tax

Property tax

Excise tax

Mineral tax

Land tax

Water and embankment tax

Tax on extraction of forestproductsTax on fisheries

Municipaloffices

GAD

Mag

we

Rak

hin

e

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

11B - % share of total


15

B.2 Fiscal transfers

28. Background: Fiscal transfers from the Union play a vital role in Myanmar’s fiscal decentralization, as revenue

generation remains centralized. Transfers can permit benefits of decentralization while at the same time mitigating

some of the potential adverse effects (World Bank, 2007, p.32). They are a necessary complement to decentralization

and will gain in importance as the country’s decentralization moves forward.

29. The number of fiscal transfers from the Union government has decreased, while the total Kyat amount

transferred has been consistently increasing. Before 2016-17, at least some SNGs received numerous small transfers

from the Union government. Examples of these transfers are: “Township Development and Management Fund”

(2015-16), “Regional Development and Poverty Alleviation Fund” (2012-2016), and various other special project

funds.15

30. Intergovernmental transfers in Myanmar amount to around 2 percent of GDP.16 An increase from 0.6

percent of GDP in 2011-12 (World Bank, 2015b). They represent 10 percent of Union government revenues in 2017-

18 BE and finance 79 percent of all SNG expenditure. Current transfers to state and regions can be characterized

under 3 components; (i) the general-purpose “deficit” grant transfer (ii) constituency development fund/grant (CDF)

and (iii) tax sharing transfer (Figure 12A and 12B).17

31. Policy objective: Intergovernmental transfers have helped support macro-fiscal objectives, closing the vertical

fiscal gap while maintaining a degree of resource equity across the country. Going forward, they may yet offer a deeper

scope as a policy tool for the Union Government. For example there are currently no conditional or performance

based grants in Myanmar’s intergovernmental transfers. It’s important that policy changes and introduction of new

transfer mechanisms aim to uphold the rules-based approach and the emphasis on improving the horizontal resource

equity.

15 For a more detailed example, see, Bago Budget Brief No.2, Figure 13 - https://rimyanmar.org/sites/rimyanmar.org/files/publication_docs/budget

_brief_series _-_bago_budget_brief_no_2_2017_english_0.pdf 16 GDP: IMF article IV report, 2017 17 General-purpose grant transfers from the Union to state/region governments were devised as deficit financing.

Box 2: Excise revenue - are states and regions missing out?

“… although Schedule 5 assigns ‘Excise Revenue’ to state/region government, what is actually collected by GAD appears

simply to be the annual license fee for running liquor-making or retail businesses. Yet, ‘excise revenue’ in English (“YitMyo”

in Myanmar) denotes a fiscal levy which has a much broader base than just license fees, whether judged by international

convention, or even apparently by Myanmar’s own legal precedents.

Under the Union Tax Law of 2016, on behalf of the Union government, the Internal Revenue Department collects ‘special

goods taxes’ on alcohol and tobacco which, based on international convention and on Myanmar’s own legislative precedent,

would appear to be ‘excise revenues’, and hence to be under the authority of state/region government to collect, rather than

the Union government. There may be some legal or regulatory reason why such taxes are not being, or cannot be, collected

by states/regions, and instead are collected by the Union government, due, for example, to Union laws or regulations not

having been re-aligned to Schedule 5 provisions; or this may simply be an oversight needing to be rectified; or it is just a

mistranslation of terms between Myanmar and English.”

Source: Abstract from Shotton at al., 2016, pp.36-37


16

32. Institutional arrangements: Intergovernmental transfer system in Myanmar is simple and more transparent,

than previously and relative to other countries of similar income level. Table 3 summarizes the institutional landscape

underpinning legal basis, design and approval elements of the 3 transfers. It suggests some institutional fragmentation

in the design of each individual component.

Table 3: Fiscal transfers and institutional arrangements

Intergovernmental fiscal transfer - Institutional arrangements

Transfer Legal basis Design Approval

General-purpose grant /a 2008 Constitution (Article 230: c,d)

Intergovernmental Fiscal

Relations Division within the

Budget Department (MOPF)

Approved by the Finance

Commission and the Union

legislature as part of the annual

budget law, no separate bill

Union tax

revenue

sharing

Special Commodity (Goods) Tax

(net of tax on imported goods)

Income Tax Law Amendment

(Article 24) The Union Cabinet approves allocation criteria in accordance

with the law. It sets and approves sharing rates with input from

the Internal Revenue Department (IRD) under MOPF.

Individual Income Tax Commercial Tax Law (Article 31)

Stamp Duty tax Special Goods Tax Law (Article 33,

Section A)

Constituency Development Fund Pyidaungsu Hluttaw Law No

9/2014

Union Hluttaw with

recommendations from the

Union Hluttaw office (see

chapter 3 of the relevant law)

Approved by the Finance

Commission and the Union

legislature as part of the annual

budget law

Notes: /a sometimes referred to as the MTFF or “deficit” grant

33. The general-purpose grant is the largest fiscal transfer to states and regions. Outside of Yangon and

Mandalay, it is the largest source of revenue and is central in the current intergovernmental fiscal system. In 2017-18

it accounted for 87 percent of all transfers.

34. As part of the medium-term fiscal framework (MTFF) reforms, the rules-based mechanism has been

introduced into the general-purpose grant starting in 2015-16. These changes have been a significant and positive

step. The transfer pool is now linked to macro indicators and a formula was introduced to guide the horizontal

allocation of funds.18 Lack of transparency in how allocation decisions are made impedes sound budget planning.

18 Application of the formula requires a deeper discussion. As it stands, the general grant transfer formula is based on the 6 equally weighted indicators: 1)

State/Region Population 2) State/Region Poverty Index 3) State/Region per Capita GDP 4) State/Region Land Area 5) State/Region Urban Population 6) Per

Capita Tax Collection in State/Region.

Figure 12: Fiscal transfers from the Union Government By type, 2017-18 BE

Source: MoPF; RI staff estimates Note: BE = Budget Estimate (enacted by legislature)

0

500,000

1,000,000

1,500,000

2,000,000

General-purposegrant

ConstituencyDevelopment

Fund (CDF)

Union Taxsharng

12A - Total, million Kyat

0

50,000

100,000

150,000

200,000

250,000

300,000

12B - Kyat per capita by states/regions

Union Tax sharng Constituency Development Fund (CDF) General-purpose grant


17

These more transparent horizontal allocation rules set a harder budget constraint and provide SNGs with a clearer

expectation of available resources earlier in the budget cycle. The ceiling for the 2018-19 FY starting in October 2018

was formally communicated in May of the same year. An improvement from 2012-13, when the ceiling, for at least

some SNGs, was announced just days before the start of the new fiscal year (Myanmar PEFA Assessment, 2012).

35. Tax sharing accounts for around 9 percent of all transfers in 2017-18 budget estimate but could play a

more important role as tax administration improves. However, this could be politically contentious, as Myanmar

increases its tax collection and brings a larger share of extractive resource industry into the tax system. In 2016-17

commercial and special goods taxes were added to individual income and stamp duty taxes which were being shared

since 2012. These 4 taxes are shared to a lower level of government – and are all collected by the Internal Revenue

Department (IRD) of the Union Ministry of Planning and Finance (MOPF) (Table 4). Although, it is possible that

sharing of the stamp duty tax has been discontinued in 2017-18 FY. Commercial and special goods taxes are the largest

by volume, accounting for over 80 percent of all tax sharing amount in 2016-17 FY (Figure 13). Income tax and

commercial tax (which are among the shared taxes) also make the majority of Union tax collection, at 3.2 and 2.6

percent of GDP in 2016-17, respectively (World Bank, 2017). With continued progress in tax administration reform,

collection of these taxes will likely grow – offering more sub-national resources under the current tax sharing scheme.19

36. A likely change to distribution mechanism of shared taxes is shifting allocation from Yangon region

to other SNGs. At the outset tax revenue was shared by origin of accrual, according to the state/region where it was

collected. In 2016-17 fiscal year, nearly 87 percent of the tax-sharing pool went to Yangon. This was due to greater

commercial activity in the region and better-developed administration (e.g. Large Tax Payer Offices (LTO)). This

creates an environment where a number of large companies pay their taxes in Yangon, not necessarily the state or

region of operation. In fact, according to government officials and the letter to the President’s Office, the change in

tax sharing allocation mechanism for 2017-18 FY was driven by equity considerations and concerns that Yangon LTO

collects taxes from businesses and the largest State Owned Enterprises (SOEs) which operate across the country.20

The change in policy reallocates 15 percent of commercial and special goods tax receipts from the SOEs at Yangon’s

LTO across states/regions in accordance with the formula used for the general-purpose grant transfer. As a result,

Yangon’s share of the Union shared taxes has decreased from nearly 87 percent in 2016-17 FY to 56 percent in 2017-

18 FY (Figure 14A and 14B).

19 For more on progress and reforms in Myanmar’s tax administration see, “Tax system efficiency” chapter in World Bank, 2017. 20 Interview, Union Ministry of Planning and Finance (May 2017).

0%

25%

50%

75%

100%

2016-17

Income tax

Stamp duty

Special goods tax

Commercial tax


Figure 13: Tax revenue sharingBy tax item, % of total shared

Table 4: Revenue Sharing Arrangements

Basis of allocation in

Tax source 2016-17 FY 2017-18 FY (supposed change)

Commercial Tax (net of tax on

imported goods)

15% by state/region

of collection

15% of tax receipts from the largest

SOEs at Yangon’s LTO are

redistributed across states/regions

according to the general grant

transfer formula.

Non-SEE, cooperatives and private

sector tax receipts are allocated by

origin like in the previous year

Special Commodity (Goods) Tax

(net of tax on imported goods)

Individual Income Tax

5% by state/region

(possibly township)

of collection

No changes

Stamp Duty Tax

2% by state/region

(possibly township)

of collection

N/A – possibly removed

Source: Adapted from Shotton et al. (2016), interviews with the Internal Revenue Department

Notes: LTO = Large Taxpayers Office; SEE = State Economic Enterprise


18

37. Constituency Development Fund makes up the smallest element of the intergovernmental fiscal

system – 1.7 percent of all transfers. Starting in 2013-14 FY, 33 billion Kyat has been equally shared among 330

townships, at 100 million Kyat each. These funds fall under the authority of Union Hluttaw Members and recorded

on state/region budgets. They tend to be used for small-scale infrastructure, and potentially supplement Union

expenditure on education and health – sectors not formally recognized on SNG budgets. While it doesn’t make a large

share of local finances, the equal amount allocation per township makes it highly inequitable on per capita basis.

38. MTFF reforms to fiscal transfers in 2014-15 FY have set a harder constraint to the size of the transfer

pool but had a smaller impact on distributional outcomes (Figure 15A). The central transfers used to be

calculated based almost exclusively on estimates of the state/regional level operating deficits. The change imposes

greater stability to fiscal resource flowing from Union to SNG budgets, hence reducing potential fiscal risk to the

central budget stemming from state/region deficits. It appears the formula is applied on the margin, to the new funds

(tied to macro indicators) added to the transfer pool, thus resulting in horizontal allocation based on both historical

distribution and application of the formula. Post-formula distributional outcomes haven’t changed by a large margin,

although does appear to have reduced variation from year-to-year (Figure 15B). In fact, this marginal distributional

impact may reflect explicit pragmatism in design, aiming to institutionalize rules without creating significant distortions

and big “losers” from the formula-induced reallocation.

Figure 15: General-purpose grant to states/regions

Figure 14: Union tax sharing

Source: MoPF; RI staff estimates. Note: Figures exclude sharing of stamp duty tax; BE = Budget Estimate (enacted by a legislature)

0%

25%

50%

75%

100%

2016-17 BE 2017-18 BE

Yangon Others

14A - Yangon & other SNGs, % of total pool

0

6,000

12,000

18,000

24,000

30,000

0

1,000

2,000

3,000

4,000

5,000

14B - Kyat per capita, by state/region

2016-17 BE 2017-18 BE

Source: MoPF: RI staff estimates

0

400,000

800,000

1,200,000

1,600,000

2,000,000

20

12/1

3

20

13/1

4

20

14/1

5

20

15/1

6

20

16/1

7

20

17/1

8

15A - Total amount, million Kyat2012-2018

0%

2%

4%

6%

8%

10%

12%

14%

16%

18%

15B - % share of total, by state/region

2013/14 2014/15 2015/16 2016/17 2017/18

post-formulapre-formula


19

39. General-purpose grant is at times adjusted. In recent years the total actual (A) transfer amount has been

revised down from the budget estimates (BE) and shares allocated to states/regions re-adjusted (Figure 16). It’s not

entirely clear how these changes are made. They’re likely aimed at correcting disparities from other sources of revenue,

most notably those created by Union tax-revenue sharing. Yangon region has consistently seen its total amount revised

down by at least 40 percent, reflecting its greater revenue performance relative to other SNGs. Moreover, as per the

2017 Myanmar PFM regulations (Article 66), the actual amounts could also be adjusted and allocated in response to

actual cash balances of SNGs.21,22 The general-purpose grant is allocated through 2 payments, the first equal to exactly

half of the total budgeted transfer amount in the enacted budget. The second payment is distributed closer to the end

of the fiscal year and adjusted according to sub-national cash balance in the previous fiscal year.23

40. Equity: The general-purpose grant does correct some disparities across regions and states, however questions

have been raised about its equity implications over time (Shotton et al., 2016) (Figure 17A). Resource equity has

worsened following the large increases in total transfers after 2012-13 FY – with the ratio of per capita resources

between that of the highest and the lowest SNG growing from 4.8 in 2012-13 to 12.4 in 2017-18 (with 2 notable

outliers in Chin and Kayah states). Overall, equity disparities are partially offset by the general-purpose grant, which

lowers the equity ratio – 21.2 and 12.4 equity ratio without and with the general grant in 2016-17 FY, respectively

(Table 5). The general-purpose grant transfer is nearly 8 times the size of tax-sharing transfers, and provides a fiscal

buffer to correct equity disparities from own-source revenue and other transfers. But current equity levels may be

challenged if other less equitable transfers grow larger in size.

21 Myanmar PFM Regulations (35/2017) https://www.mopf.gov.mm/sites/default/files/upload_pdf/2017/08/Financial%20Rules%20and%20Regulation.pdf 22 Myanmar PFM regulations. Chapter 7, Section 66. 23 For example, if state/region has a 3 million Kyat surplus at the end of a given fiscal year, this amount would be subtracted from the formula generated

amount the following fiscal year.

0%

20%

40%

60%

80%

100%

120%

140%

2013-14 2014-15 2015-16 2016-17

Figure 16: General-purpose grant deviationsActual transfers (A) as % share of budgeted (BE), 2012-2017

Kachin Kayah Kayin Chin Sagaing

Thanintaryi Bago Maggway Mandalay Mon

Rakkhine Yangon Shan Ayeyarwaddy

Source: MoPF; RI staff estimates. Notes: BE = Budget Estimate (enacted by legislature); A = Actual (internally audited figures)

Rakhine

Magwe

https://www.mopf.gov.mm/sites/default/files/upload_pdf/2017/08/Financial%20Rules%20and%20Regulation.pdf


20

41. Formula specific comments: The general-purpose grant uses a formula with 6 equally weighted indicators: 3

as proxies for expenditure needs and 3 as proxies of fiscal constraint (in other words, reflecting ability to generate own

revenues).24 Some of the common comments with regards to the formula include: i) GDP per capita as an imprecise

proxy for SNG revenue potential – since local GDP is poorly correlated with local tax performance; ii) Poverty Index

values don’t appear to be weighted by relative population – as a result SNGs with similar poverty values but different

sized population (hence different resource demands for addressing the poor) would receive the same transfer amount

associated with the variable. iii) Per capita tax collection variable could potentially act as a disincentive for improved

tax collection – with above average performers receiving less of the transfer amount from the 1/6th of the pool

associated with the variable; iv) the fiscal constraint variables take 3/6th (50%) weight of the pool – possibly excessive

fraction, since nationally own-source revenues account for a small share SNG revenues.

42. Transparency and consensus: Technical considerations are an important element of intergovernmental

transfer design and have an impact on resource equity, fiscal risks, as well as incentives over tax collection. No less

important are institutions underpinning the design and transparency of the system. International experience suggests

that allocation of fiscal resources across SNGs can become highly political, sometimes at expense of regional equity

and fiscal sustainability. Trade-offs from future policy changes to fiscal transfers have to reconcile with political reality,

status quos and divergent fiscal demands across the country. Long-term sustainability of the system would demand

building mechanisms to communicate trade-offs across states/regions and balancing needs in an open manner. These

mechanisms are important to consider as changes to the system are made. (for discussion on institutional arrangements

of fiscal transfers see, World Bank, 2007, chapter 10)

24 The general-purpose grant transfer formula is based on the 6 equally weighted indicators: 1) State/Region Population 2) State/Region Poverty Index 3)

State/Region per Capita GDP 4) State/Region Land Area 5) State/Region Urban Population 6) Per Capita Tax Collection in State/Region

Figure 17: Resource equity ratios Total available fiscal resources, Kyat per capita by state/region

Source: MoPF; RI staff estimates Notes: Total available resources include all of the budgeted own-sourced revenue and all identifiable transfers. BE = Budget Estimate (enacted by legislature)

0

50,000

100,000

150,000

200,000

250,000

300,000

2012-13 2017-18

2012-2018

Kachin Kayah Kayin ChinSagaing Thanintaryi Bago MaggwayMandalay Mon Rakkhine YangonShan Ayeyarwaddy

RakhineMagwe

Table 5 – With and without the general-purpose grant, 2016-17 BE

With Without

Chin 278,829 5,774

Kayah 184,048 8,077

Thanintaryi 108,396 5,165

Kachin 100,610 7,911

Yangon 75,278 69,727

Rakhine 70,665 3,766

Kayin 50,235 3,287

Shan 43,540 6,385

Magwe 40,515 4,385

Mon 38,960 4,416

Sagaing 37,779 4,822

Mandalay 37,682 20,291

Bago 30,866 4,905

Ayeyarwaddy 22,430 3,642

Max:Min ratio

12.4 21.2


21

C. Expenditure

43. State and region expenditure has nearly tripled since their first enacted budgets in 2012-13 (Figure

18).25 Annual expenditure growth across 14 states/regions averaged 23 percent, with an overall increase of 286 percent

– from 864 million Kyat to 2,475 million Kyat between 2012-13 and 2017-18 fiscal years. Increases in expenditure

have been largely driven by increases in fiscal transfers from the Union Government (Figure 19).

44. The economic composition of sub-national budget expenditure varies significantly across states and

regions. Overall, capital spending is large as a ratio of SNG budgets – with all states/regions spending at least 40

percent of their budgets on capital investments, and as high as 78 percent in Chin state (Figure 20). Maintenance costs

vary between 6 percent of the budget in Rakhine state and 45 percent in Tanintharyi region.

25 State/region expenditure mandate is broadly guided by the Schedule 2 of the constitution

0

100,000

200,000

300,000

400,000

500,000

600,000

700,000

800,000

0

50,000

100,000

150,000

200,000

250,000

300,000

350,000

400,000

2012-13 2017-18 2012-13 2017-18

Figure 18: Total SNG expenditureBy state/region, million Kyat 2012-2018

Kachin Kayin Chin Sagaing Thanintaryi

Kayah Bago Maggway Mandalay Mon

Rakkhine Shan Ayeyarwaddy Yangon

Source: MoPF, RI staff estimates

Yangon

Shan

Kayah

Rakhine

Magwe

-100%

0%

100%

200%

300%

400%

-100% 0% 100% 200% 300% 400% 500%

Yea

r-o

n-y

ear

exp

end

itu

re

gro

wth

Year-on-year growth ingeneral-purpose grant transfer

Figure 19: Growth in expenditure and fiscal transfersAcross 14 states/regions, 2012-2018

0%

20%

40%

60%

80%

100%

Figure 20: Expenditure sharesBy economic categories, 2016-17 RE

Capital Salary Maintenance Other

Source: MoPF; RI staff estimates. Note: RE = Revised Estimate (revised figures during fiscal year)


22

45. Capital expenditure has grown as a share of total sub-national spending. The share of budgeted capital

expenditure increased from 36 percent in 2013-14 to nearly 55 percent in 2016-17. The growth in capital expenditure

may reflect states and regions increasingly prioritizing infrastructure development (Figure 21).26 Yangon region is the

only SNG to have reduced its share of budgeted capital expenditure during the same period, from 58 percent to 45

percent between 2012-13 and 2016-17 fiscal years.

46. Increases in capital expenditure, as a higher proportion of SNG budgets, are more pronounced in

Myanmar’s states, than in the regions (Figure 22).27 This may reflect the need for greater infrastructure spending,

such as on roads and bridges, in the traditionally poorer border areas. The capital share of budgets for states has grown

from 13 to 65 percent between 2012-13 and 2016-17 FYs (Figure 22B). For regions, this figure has grown from 41 to

49 percent during the same period. Regions still account for the larger share (55 percent) of overall capital expenditure

across all SNGs in 2016-17 FY, although a significantly lower figure than 88 percent share in 2012-13 FY (Figure 22C).

Figure 22: Capital expenditure


Notes: States are Chin, Kachin, Kayah, Kayin, Mon, Rakhine, and Shan. Regions are Ayeyarwaddy, Bago, Magwe, Mandalay, Sagaing, Tanintharyi and Yangon.

RE = revised estimate (revised figures during fiscal year); A = Actual (internally audited figures)

26 For more on capital expenditure in states/regions see, World Bank, 2017, chapter 2. 27 It is important to note that the variation between capital and current expenditure ratios could be at least partly driven by ambiguity in accounting

definitions – see Shotton et al., 2016, p.62.

0%10%20%30%40%50%60%70%80%90%

Figure 21: SNG capital expenditure% share of total, 2012-13 & 2016-17 2012-13 2016-17


0%

10%

20%

30%

40%

50%

60%

70%

22B- % of total budgets

states vs. regions, 2012-2017

0%

20%

40%

60%

80%

100%

22C - % of all capital budgets

states vs. regions, 2012-2017

22A - Share of budget

across SNGs, 2016-17 (RE)


23

47. Budgeted figures may not reflect the true pattern of capital expenditure. The growth may be, at least

partially, attributed to accounting norms, and changes in budget classification changes. In practice, classifications

across time and ministries are not necessarily consistent with each other. It’s not certain there is guidance and

definitions of capital expenditure categories (for example, in the form of a manual). On the other hand, capital

expenditure may also be underreported. It was suggested that capital investments are broken down into smaller

projects and sometimes classified under current accounts for various reasons.28 These inconsistencies in classification

blur the picture of overall sub-national investments and may be a source of potential risk.

48. Road spending dominates sub-national spending (Figure 23). Most of it executed by and budgeted

under Department of Highways (DOH) within the Ministry of Construction (MOC). A share of municipal budgets is

also spent on roads (Winter and Nandar, 2016). More recently, rural road expenditure by the newly established

Department of Rural Road Development (also under MOC) has also shifted to sub-national budgets (for more on

rural roads expenditure see, Valley et al., 2018). This expanded expenditure mandate for rural roads has not been

supplemented by additional funds to sub-national budgets.

49. State Economic Enterprises (SEE): SEEs used to make a significant share of SNG budgets, predominantly

through road investments under Public Works – formally under MOC the body was responsible for development,

maintenance and operation of the entire trunk road network and accounted for 54 percent of all sub-national spending

in 2013-14 (Dickenson-Jones et al., 2015, p.34). The government has since reformed Public Works into government

departments, with DOH being the largest.29 Currently, the only sizable SEE to appear on SNG budgets is Electricity

Supply Enterprise (or Electricity Distribution Work) under the Ministry of Electrical Power and Energy, which covers

as high as 19 percent of the 2017-18 budget estimate in Tanintharyi.

50. Expenditure by cabinet offices: Sub-national government bodies account for about 21 percent of

expenditure in a sample of 9 states/regions – nearly all of it budgeted under the sub-national cabinet offices. Available

budget formats provide aggregate figures only, making it hard to identify what the money is spent on. Although,

budgets for Bago and Kayin suggest that most of it is allotted to irrigation and agriculture projects (Renaissance

Institute, 2017-2018).30 Interviews with local officials suggest that the new expenditure on rural roads is also recorded

under the cabinet budgets. It’s unclear why this expenditure is not recorded under the budgets of their relevant

28 A remark from Pyithu Hluttaw Member of Parliament suggested this is done to avoid the legally mandated oversight of capital investments by Hluttaw

members. It was also suggested that smaller value projects are more likely to get approved 29 Under Ministry Resolution #21/2015 issued on 31 March 2015, Public Works split into separate government departments 30 57 and 28 percent of Ayeyarwaddy’s cabinet office’s capital expenditure during the 2018 interim budget was spent on irrigation and electricity projects,

respectively

0%10%20%30%40%50%60%70%80%90%

100%

Figure 23: Sub-national expenditureBy administrative units, % of total across 9 state/region budgets in 2017-18 BE

Other


Ministry of Electrical Power and Energy

Municipal

Sub-national government bodies

Ministry of Construction

Source: OMI; RI staff estimates Note: Other includes: 1) Ministry of Agriculture, Livestock and Irrigation; 2) Ministry of Health and Sports; 3) Ministry of Natural Resources and Environmental Conservation; 4) Ministry of Planning and Finance, and few other small departments; Sub-national government bodies includes: SNG Cabinet Office, SNG Parliament, Court, Attorney General Office and Auditor Office; Municipal units include Development Affairs Organizations (DAOs) and Mandalay City Development Committee (MCDC); BE = Budget Estimate (enacted by legislature)


24

executing departments. Such practices create challenges to oversight, transparency, and accountability of expenditure.

It also blurs the functional perspective of expenditure, making it hard to more accurately estimate what share of the

budget is spent on roads, electricity or agriculture. Anecdotal evidence also suggests that cabinet offices get “requests”

from Union line ministries to budget for specific items and projects, potentially those not formally recognized as SNG

mandate under schedule 2 of the constitution. It’s not clear whether this reflects a pattern or a few isolated instances.

51. Most of the capital expenditure is budgeted under MOC, followed by sub-national government bodies

and Municipal offices (Figure 24). The former’s dominance of the capital budgets across a sample of 9

states/regions is not surprising given the prominence of road spending in local budgets. Interestingly, nearly a quarter

of total capital budgets in the same sample is recorded under the sub-national cabinet offices, which spend nearly half

of their budgets on capital expenses (Figure 25).

Source: OMI, MoPF; RI staff estimates.

Notes: Across 9 states/regions: Kachin, Kayin, Sagaing, Tanintharyi, Bago, Magwe, Mandalay, Mon, and Ayeyarwaddy; BE = Budget Estimate (enacted by a legislature).

52. Current expenditure breakdown and trends: Total current expenditure across SNGs has decreased from

1,340 billion Kyat to around 950 billion Kyat between 2014-15 and 2017-18 fiscal years. Most of this decline can be

attributed to a decline in “purchases of good and services” (Figure 26). Around 40 percent of all sub-national current

expenditure in 2017-18 FY is budgeted as “maintenance”, followed by 29 percent as “salary payments” and 15 percent

as “purchases of goods and services”.

17%

22%

42%

12%

4% 3%

Municipal

Sub-national government bodies

Min. Construction

Min. Electrical Power and Energy

Min. Home Affairs

Other

Figure 24: Capital expenditure breakdownBy ministry as % of total capex, 2017-18 BE

0%

20%

40%

60%

80%

100%

Municipal Sub-nationalgovernment

bodies

Min.Construction

Min.Electrical

Power andEnergy

Min. HomeAffairs

Other

Current Capital

Figure 25: Expenditure breakdownBy type as % of total Ministry expenditure, 2017-18 BE

0

10,000

20,000

30,000

40,000

50,000

60,000

70,000

0

100,000

200,000

300,000

400,000

500,000

600,000

700,000

2014-15(A) 2015-16 (PA) 2016-17 (RE) 2017-18 (BE) 2014-15(A) 2015-16 (PA) 2016-17 (RE) 2017-18 (BE)

Figure 26: Sub-national current expenditureTotal of 14 states/regions in millions of Kyat, by economic classification, 2014-2018

21 Compensation of employees 22 Purchases of goods and services Maintenance

27 Social benefits 24 Interest payment Taxes paid to Union Budget

Contribution paid to Union Budget Other expenses

Source: MOPF; RI staff estimatesNotes: Numbers in the legend approximate IMF GFS codes: Purchases of goods and services include expenses on gasoline, lubricant, engine oil, travel and hosting expenses, as well as SOE related budget items like operational expenses, administrative and research expenses, and sale and distributional expenses. Social benefits include employment related social benefits like educational trainings, social assistance, and social security. Taxes paid to Union are related to commercial and income taxes of SOEs.Other expenses include other unclassified expenditure on land projects, irrigation, dam, lake, drainage, and other agricultural water supply works. BE = Budget Estimate (enacted by legislature); RE = Revised Estimate (revised figures during fiscal year); PA = Provisional Actual; A = Actual (internally audited figures).


25

53. Expense accounting: The boundary between maintenance, purchases of goods and the acquisition of fixed

nonfinancial assets (capital expense) may be blurry in practice. Experience from working with budgets across the

country suggests that classifications across time and ministries are not always consistent with each other. For example,

machinery and equipment forming an integral part of a building could be included as either maintenance or capital

expense. Equally, goods acquired for use as fixed assets or valuables, for use in capital formation under a certain Kyat

threshold could be classified under current accounts, rather than capital.


26

D. Institutions and management

D.1 Budget execution and credibility

54. Effective public service delivery and successful investment are more likely when ministries and their

departments and enterprises spend the amounts approved for them in the budget. Budget execution, particularly of

the capital expenses, is cited as the main reason for shifting the budget calendar year, allowing for more time within

the budget cycle before the rainy monsoon season. While weather-related constraints could slow down the execution

of investment projects, there could potentially be other reasons for deviations from the approved budget. Poor

execution of the budget could also be linked to cash management issues, procurement practices, lack of fiscal

responsibility, poor quality of data and others.

55. Revenue out-turns have deviated far from budgeted amounts over the 5 year period (Figure 27).

Average deviations (in either direction) from the budgeted amounts, across 14 SNGs, have halved between 2012-13

and 2015-16 FYs, from 30 to 15 percent. They have since increased to around 27 percent in 2016-17. The direction

of misestimation has changed over the years, with at least 4 states/regions over-estimating revenue between 2012-13

and 2014-15 FYs (7 in 2014-15). During the more recent, 2015-16 and 2016-17, fiscal years all states/regions have

been consistently underestimating their revenues.

56. Initial budgetary targets for revenue are calculated by each of the revenue collecting ministries. No

routine revenue forecasts beyond the annual targets are performed. Further examination is required to explain these

large misestimations. It could be partially attributed to incentives of reaching internal ministry revenue targets and

setting of low targets in the first place.

57. Execution rates have improved with MTFF reforms. Average deviations from budgeted expenditure

amounts have declined since 2012-13 FY (Figure 28). The average weighted deviation across states/regions was around

18 percent in 2012-13, with five states/regions underspending relative to their budgeted figures. The picture has

improved with a shift towards a more rules-based intergovernmental fiscal system, with average weighted deviations

declining to 3 and 8 percent in 2015-16 and 2016-17, respectively. According to revised estimates, in 2016-17 all

states/regions with the exception of Yangon have overspent within 10 percent of the approved budget. Yangon’s

spending of 50 percent over the approved budget in 2016-17 has been compensated by, and possibly influenced by,

unbudgeted revenue surpluses the same year (Figures 27 and 28).

0%

50%

100%

150%

200%

2012-13 (A) 2013-14 (A) 2014-15 (PA) 2015-16 (RE) 2016-17 (RE)

Figure 27: Own-revenue estimation and budget credibilitySub-national revenue as % share of budgeted (BE) - excluding the general-purpose

grant transfers and Union shared taxes, 2012-2017




Source: MoPF; RI staff estimates. Notes: BE = Budget Estimate (enacted by legislature); RE = Revised Estimate (revised figure during the fiscal year); PA = Provisional Actual; A = Actual (internally audited figures).

Average revenue out-turn

Magwe

Rakhine


27

58. There is scope for more analysis of budget credibility and expenditure controls. Aggregate budget execution

figures may hide deviations across ministries and departments. Moreover, while the aggregate may appear close to

total budgeted amounts, the actual target and cost of expenditure could potentially differ from what is in the budget

laws approved by the legislature. Although expenditure controls are not explored in this brief, they are an important

element of technical support to the government. Medium-term plans and annual budgets may be less meaningful if

expenditure cannot be controlled during execution. A lack of effective expenditure controls may undermine fiscal

discipline as well as the trust in a government’s stewardship of public resources.

D.2 Cash management and change to the fiscal calendar

59. Cash management refers to having the right amount of money at the right place and the right time for the

government to meet its obligations. It naturally relates to the budget execution, if cash is not there to pay suppliers

and contracts projects may be delayed, face cost overruns or even cancellation. In the context of Myanmar SNGs, the

challenge may arise when there is a timing mismatch between revenue (tax and non-tax revenue, transfers, and

potentially borrowing) and cash outflows in the form of salaries, operational expenditure, acquisition of assets, and

potentially debt repayments. This subsection provides an introductory overview of cash management.

60. SNG Fund Accounts: State/Region funds are consolidated at the State/Region Fund Account (FA) within

the Myanma Economic Bank (MEB). According to Myanmar’s PFM regulations, the Myanmar Central Bank is

responsible for providing banking services to the government and it devolves some operational responsibility to the

MEB, which uses its network of branches to manage accounts of the general government (Article 187). The

state/region governments are in turn responsible for management of these accounts in accordance with the respective

enacted budget law (PFM regulations, Article 11). In practice, the state/region budget departments provide a degree

of oversight of cash flows through its internal reports. It also contributes to cash management by facilitating the

movement of the Union transfers. Every department that appears on a sub-national budget has its own account at

MEB. The exact nature of the relationship between line ministry departments and MEB, sub-national cash flows, cash

management and forecasting is not explored in this report and may require further examination.

61. Cash reporting – budget department: Each department under the state/region FA submits monthly cash

reports to relevant state/region budget departments. The latter aggregates these reports and sends its own monthly

cash flow reports to the Union Treasury department, to the relevant state/region cabinet, minister of finance and

auditor general’s office. These reports include opening and closing balances by administrative units and type of

expenditure (e.g. current, capital, finance, interest, and loss).

0%

20%

40%

60%80%

100%120%

140%

160%

180%

2012-13 (A) 2013-14 (A) 2014-15 (PA) 2015-16 (RE) 2016-17 (RE)

Figure 28: Budget execution ratesSub-national expenditure as % share of budgeted amount (BE), 2012-2017




Source: MoPF; RI staff estimates. Notes: BE = Budget Estimate (enacted by legislature); RE = Revised Estimate (revised figure during the fiscal year); PA = Provisional Actual; A = Actual (internally audited figures).

Rakhine

Magwe


28

62. Cash reporting – MEB: MEB provides its own monthly reports to the Central Bank of Myanmar and the

relevant state/region budget department. These reports include a number of statements: monthly cash reports on

payments, opening and closing balances for State Economic Enterprises (SEE), development committee (municipal

office), ministries and departments, and other account.31

63. Cash surpluses: Any cash surpluses at the state/region fund accounts are recalled to the so-called General

Reserve Fund (GRF) at the end of the fiscal year. These surpluses stay in local state and region fund account to be

used during the following budget cycle. As mentioned earlier in section B2, cash balances also interact with the general-

grant transfer. Surplus amounts in GRF at the end of the fiscal year are subtracted from the total estimated grant

amount for the following fiscal year.

64. Cash flows and budget execution: A brief look at Kayin’s cash inflows and outflows in 2016-17 shows that

nearly half of the total budget was executed in the final quarter (Figure 29). The most significant inflows happen twice

a year, as part of the general-purpose grant transfer. Given the timing of transfer disbursement during the 2016-17

fiscal year, nearly half of the budget resources were made available 3 months before the end of the fiscal year (Figure

30). It’s hard to identify whether budget execution is significantly affected by late disbursement of the fiscal transfers

from the Union Government or is driven by other considerations.

Source: Kayin Budget Department; RI staff estimates

65. Fiscal calendars: Since cash management relates to the timing of inflows and outflows of cash it also interacts

with budget and tax calendars. Tax and budget calendars in Myanmar used to be identical, from April to March.

Starting in October 2018 Myanmar will shift its budget calendar by 6 months, to begin October and terminate in

September. Currently, the tax calendar (outside of SEEs) is expected to remain the same, April to September (Table

6). Different calendars are not uncommon around the world. However, changes to either require some consideration.

Table 6: Budget and tax calendars

April July October January April July October January April July October

Tax calendar 2018-19 Tax calendar 2019-20

Interim budget Budget calendar 2018-19 Budget calendar 2019-2020

66. Tax calendar: If tax policy changes, under the new budget calendar, have to be approved alongside the budget

law (October), the delayed start of the following tax calendar may create incentives for taxpayers to shift their payments

from one tax period to another, and hence challenging the government’s cash forecasting and balance. On the other

31 For statements in Bago, other account included balances for the Department of Rural Road Development (DRRD) and the Regional Government Office

(cabinet)

7%

18%

26%

48 %

First Quarter Second Quarter

Third Quarter Fourth Quarter

Figure 29: Budget executionBy quarter as a share of total, Kayin budget, 2016-17

0

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

Figure 30: Monthly revenue and expenditure Kayin State, 2016-17, million of Kyat

Union grant transfer Other revenue Expenditure


29

hand, aligning the tax and budget laws may impose a significant burden on the tax administration in shifting its

processes and educating the broader public about changes to tax payment dates. If poorly managed, this could result

in loss of tax revenue. Given SNGs low reliance on local tax sources, this is more likely to affect them in indirect ways,

for example, through Union tax sharing and possibly other transfers.

D.3 Fiscal reporting and transparency

67. Fiscal transparency is a critical element of effective fiscal management. It relates to comprehensiveness, clarity,

reliability, timeliness, and relevance of public reporting on the past, present, and future state of public finances. Fiscal

transparency is critical for effective fiscal management and accountability. It helps ensure that governments have an

accurate picture of their finances when making economic decisions, including of the costs and benefits of policy

changes and potential risks to public finances (IMF, 2018). It also provides legislatures, markets, and citizens with the

information they need to hold governments accountable.

68. Availability: Sub-national governments have made progress but significant space remains to improve public

access to key fiscal information. Myanmar’s sub-national fiscal transparency has made progress in recent years but is

still lagging behind that of the Union Government.32 Most of the sub-national annual planning and budget laws are

publicly available at Myanmar Law Information System developed by Union Attorney General Office; however, not

across the whole country and only in the middle (or later) of the budget calendar.33 There is currently no public access

to sub-national in-year budget reports, year-end financial statements, external audit reports, or information on large

contract awards. Equally important, there are no forward or policy focused public documents stating a government's

budgetary objectives, policy intentions and projections with respect to public finances.

69. Comprehensiveness: Sub-national annual budgets, as contained in the state/region budget Laws, follow

accounting classification in line with the uniform countrywide system; however, are not fully consistent with modern

classification structures. The sub-national government budget laws describe the projected values of revenue and

expenditure in aggregated headlines by administrative categories (e.g. ministries and departments). Here too, there are

opportunities for improvements. For example, by incorporating more extensive information on geographical

resources allocation by administrative units, and by functional or economic categories. On the revenue side, a more

specific breakdown of major own sources of revenue: non-tax current revenue (which makes the bulk of own-source

revenue) as well as tax receipts. Moreover, activities of several key administrative units within the SNG budgets aren’t

well covered. This includes expenditure undertaken under the state/region cabinet offices, or both

expenditure/revenue accounts of the municipal offices (especially those of YCDC/MCDC).

70. Clarity More recently, SNGs have started publishing citizen’s budgets. 6 states/regions have published at least

1 citizen’s budget since 2016.34 These citizen’s budgets present information from enacted budgets in a more accessible

language, incorporating visual elements to help non-specialist readers understand the information.

71. Reliability and relevance: Aggregate figures alone may distort the fiscal picture, and lack accuracy in

representing government operations and finances. Unreliable information may limit the public’s ability to hold

government to account. Moreover, unreliable reporting may be disadvantageous to the government itself. As noted

earlier, classifications across time and ministries are not necessarily consistent with each other. This relates not only

to disclosure of information to public but also to internal reporting, potentially blurring the overall sub-national fiscal

picture for higher level offices. Hence, obscuring the accurate picture of investments, risks and costing of future fiscal

obligations. Such weaknesses in transparent reporting can also effect the government’s ability to effectively evaluate

the quality of its expenditure, and limit the use of evidence in policy making.

32 In recent years, the Union Government published a series of budget-related documents, some of which include: the annual budget law, budget speech by

Finance Minister, Citizen’s budget, quarterly, mid-year and year-end reports on the MOPF website: www.mopf.gov.mm 33 https://www.mlis.gov.mm/ 34 Kayah State (for 2017-18), Kayin State (for 2017-18 and 2018 interim), Tanintharyi Region (for 2017-18 and 2018 interim), Bago Region (for 2017-18

and 2018 interim), Mon State (for 2017-18), Ayeyarwaddy (for 2017-18)

http://www.mopf.gov.mm/

https://www.mlis.gov.mm/


30

Box 3: Hluttaw

Parliaments’ main roles are to review and debate the government’s draft ex ante budget and to authorize spending to

implement the annual budget plan. As with the principles of transparency, here too, Hluttaw requires timely, comprehensive

and relevant information. In the past state/region Hluttaws have had very little time to approve their budgets, sometimes just

a few days (Shotton et al., 2016). The revised budget calendar for 2018-19 fiscal year may offer more time for Hluttaws to

review and approve the budgets. According to the new sub-national budget calendar, the draft budget is to be submitted to

sub-national Hluttaws between the 3rd week of April and the 1st week of May, to be approved by October 2018. This is in line

with international good practice, with submission of the draft annual budget to parliament 2–4 months in advance of the

beginning of the new fiscal year (IMF, 2010). As a result, some state/region Hluttaws have had more time to review their

budgets, ranging from around 1 to 2 months.*

The quality of information submitted for review appears to differ across the country. Budget proposals typically involve

aggregated budget data by administrative unit and type of account (e.g. current vs. capital). Project level proposals are also

submitted, although it’s unclear whether this is done systematically.** Most budget data is submitted in paper form, aggregated

across multiple physical stacks of books. Lack of access to electronic formats imposes additional time burden and challenges

the quality of review Hluttaws can provide. Arguably, it also narrows the scope for treatment of budgets as a regional policy

document, shifting a larger focus on lobbying over individual projects.

* For example, Bago budget department submitted the proposal to its regional Hluttaw on March 31, 2018 and Hluttaw’s Public Account Committee submitted its findings to the regional Hluttaw on June 4, 2018. In Mandalay, the submission of budget proposals from budget department happened on March 31, 2018. In Ayeyarwaddy, the budget department submitted the budget proposal on May 29, 2018, allowing its Hluttaw nearly a month for review. ** According to publicly available documents, it appears that Yangon Hluttaw have had little information surrounding the region’s financial revenue and expenditure.


31

E. Fiscal risks

72. Fiscal risk can be characterized as deviations of fiscal outcomes from what was expected at the time of the

budget or other forecasts. SNGs play a relatively small role in general government finances but may still face and be

a source of fiscal risk themselves. Ultimately, given their reliance on Union financing, fiscal risks faced by SNGs are

contingent liabilities to the Union Government. International experience provides multiple examples of sub-national

government bailouts by the central government. Estimates of fiscal costs associated with the realization of contingent

liabilities from sub-national governments in emerging market economies are around 4 percent of GDP, and as high

as 12 percent of GDP in Brazil in 1997 (Bevilaqua, 2002; Elva et al., 2016).

73. Budget deficits: The risk of financing unanticipated SNG deficits has been, at least to a degree, offset by

introducing a rules-based approach to intergovernmental financing. It doesn’t completely eliminate the possibility of

SNGs going over the allotted ceiling and it may be too early to tell, but early signs suggest better fiscal discipline (Figure

28). Sub-national debt does not appear to be a major source of risk, given a near absence of sub-national borrowing.

Although, a potential increase in borrowing by Yangon and Mandalay may introduce some risk.

74. Public-private partnerships: One source of fiscal risk could emanate from liabilities from public-private

partnerships (PPPs). They often entail fiscal obligations not captured in the fiscal account: for example, government

guarantees for concessionaire borrowing, minimum revenues, or exchange rate losses (Cebotari et al., 2009). As the

appetite for PPPs is starting to grow in major urban centers it's important to understand long-term fiscal implications

of these arrangements. Contractual guarantees can become a source of explicit contingent liabilities to state/region

governments.35 Limited information on contingent liabilities and future costs of investments could harm management

of long-term fiscal sustainability.

75. A large share of the highway network in Myanmar operates under the Build-Operate-Transfer (BOT)

arrangements. SNGs play an important role in financing the highway network (Valley et al., 2018), although it’s unclear

if they would also assume associated liabilities. The nature of these BOT contracts are not publicly disclosed, hence

it’s hard to tell whether they could be a source of concern.

76. Disclosure of fiscal information: Limited budget comprehensiveness and transparency could be a source of

risk in itself, by reducing the Union Government’s ability to monitor sub-national governments' fiscal position. The

budget classification system (especially of capital investments) and potential inconsistencies in internal reporting (e.g.

from SNG to Union) may restrict the central government’s ability to gather a consolidated view of fiscal risks. This

could leave the general government open to blind spots like payment arrears or, on a longer-term horizon, possibly

debt defaults. Furthermore, limited availability of information to the public limits scrutiny, opening risks for

corruption, providing potential opportunities for leakages, and inappropriate use of funds.

77. Fiscal transfers: From the perspective of sub-national governments, their heavy reliance on central transfers

could be a risk in itself. An event that prompts contraction of fiscal transfers may hurt the sub-national governments’

ability to deliver mandated services. This is not a major risk given the SNGs relatively low share of total public

expenditure. However, it may become larger if SNGs expenditure mandates are expanded without the adequate

increase in own-source revenue collection.

35 For example, substantial obligations on PPP contracts in power plants and roads became due in Indonesia, Malaysia, and Thailand during the Asian crisis.


32

F. Concluding remarks

Myanmar’s new sub-national governments are playing an increasingly important role in management of public finance,

service delivery, and ultimately economic growth. States/Regions face poor infrastructure, its citizens low access to

services, and governments remain underfunded. Successful and sustainable solutions will require a better

understanding of local systems. This report contributes to that knowledge by offering the most recent and

comprehensive outlook on sub-national public finances across Myanmar.

Revenue: Few will argue against the need to improve sub-national revenue collection. Prior research has shed light

on administration, policy and legal basis of taxes like property tax (McDonald and Hein, 2017), the importance and

context of local non-tax revenue, like municipal auction licenses (Bissinger, 2016). These provide examples where, at

least, some positive steps can be taken today. However, Myanmar’s sub-national revenue collection presents many

other sources, both formal and informal. Some are less understood but are important to broader tax reform and

maximizing the sub-national revenue space allowed within the constitution. Intergovernmental transfers play a central

role in sub-national revenue. So far, they have helped support macro-fiscal objectives, while maintaining a degree of

resource equity across the country. They may also offer a deeper scope as a policy tool while upholding the rules-

based approach and the emphasis on improving the horizontal resource equity. One of the outcomes of

decentralization is that it has fostered experimentation, including on local revenue mobilizatoin.36 Their evaluation

and knowledge-sharing can play an important role in building policy-relevant evidence and sustaining the reform

momentum.37

Expenditure: Focus on the expenditure side has primarily been over the size of sub-national budgets and total

allocation across sectors. This in itself is a step forward as budget information becomes more available and different

levels of government build more clarity in their expenditure responsibilities. Going forward, it’s important to look at

the quality and efficiency of spending. So far, this has been held back by limited access to budget data, made worse by

inconsistent and narrow reporting mechanisms. Without transparent and reliable data on budgetary inputs, evidence-

based policy-making and public dialogue underpinning the government’s most important policy document, its budget,

will be constrained. There is space in building a better understanding of sub-national budgetary inputs and their link

to socioeconomic outcomes. Better data (and public access to it) will help enrich the debate surrounding public money.

This would allow scrutiny of not only how much the government spends on agriculture or roads, but whether it spends

it effectively: that is on projects that bring the most value per spent Kyat, and if it is spent diligently.

Donor support: Donor assistance has helped make progress in public financial management at both levels of

government. This includes better forecasting, more sustainable and transparent use of funds. Their support has

facilitated, what arguably are, some of the bigger steps the government has made towards transparency in half a century.

In that spirit, there is continued demand for support of budget transparency initiatives (e.g. citizen’s budgets). It is

also important to draw more attention on less visible elements underpinning the stewardship of public money. This

includes the need for standardization of reporting formats, bringing classification practices in line with those of the

Union Government and modern standards. Crucially, this also includes efforts to build capacity within the civil service.

Budget departments are undergoing an important institutional transformation. They are moving away from a clerical-

focused agency with emphasis on aggregation of documents, to the one with more analytical and policy-oriented focus.

This transition requires a new set of skills: for example, a better grasp of technical tools in treating and communicating

of data as well as analytical tools to identify what data is useful for monitoring of quality of expenditure, fiscal risks

etc. Improved information management of public finances will also help support the new democratic instructions,

ensuring parliaments are offered more relevant and timely information to provide appropriate checks on the executive.

While such support may appear intangible, it will be crucial in supporting Myanmar’s democratic transition and helping

ensure that government funds, along with donor funds, are spent transparently and effectively.

36 An example of such experimentation includes the sale of slaughter licenses in Shan state (Bissinger, 2016). 37 Recently, horizontal learning workshops across municipal offices have been facilitated by Renaissance Institute and The Asia Foundation. Although still

in their early stage, such events could help promote evidence-based policy making across municipal governments in Myanmar.


33

References

Bevilaqua, Afonso, 2002, “State Government Bailouts in Brazil,” Research Network Working Paper No. R-441 (Washington: Inter-American Development Bank)

Bissinger, Jared. 2016. “Local Economic Governance in Myanmar”. Yangon: The Asia Foundation.

Bova Elva, Ruiz-Arranz Marta, Toscani Frederik, and Ture H. Elif, 2016. “The Fiscal Costs of Contingent Liabilities: A New Dataset” International Monetary Fund (IMF) Working Paper.

Cebotari Aliona, Davis Jeffrey, Lusinyan Lusine, Mati Amine, Mauro Paolo, Petrie Murray, and Velloso Ricardo, 2009. “Fiscal Risks: Sources, Disclosure, and Management”. International Monetary Fund (IMF)

Chan Alison, Guo Stephanie, Win Po Po Aung, 2018. “Gender mainstreaming in subnational governance: Myanmar’s fisheries and livestock sectors”, Yangon: The Asia Foundation

International Monetary Fund (IMF), 2010. “Technical Notes and Manuals: Role of the Legislature in Budget Processes”.

International Monetary Fund (IMF), 2018. “Fiscal Transparency Handbook”.

McDonald, Lachlan and Arkar Hein. 2017. "Managing the challenges of rapid urbanization: A review of the existing property tax system in Myanmar". Yangon: Renaissance Institute.

Minoletti, Paul. “Fiscal Decentralisation and National Reconciliation in Myanmar: Key Issues and Avenues for Reform”. International Growth Center, 2016 - S-53404-MYA-1

Myanmar PEFA Assessment, 2012. "Public Financial Management Performance Report".

OECD/UCLG, 2016. “Subnational Governments around the world: Structure and finance”.

Renaissance Institute, 2016-2018. “Budget Brief Series”

Shotton, Roger, Zin Wint Yee and Khin Pwint Oo. 2016. “State and Region Financing, Planning and Budgeting in Myanmar: What are the Procedures and What are the Outcomes”. Yangon: Renaissance Institute and The Asia Foundation.

Valley Ildrim, McDonald Lachlan, Hein Aung Kyaw, and Nan Sandi, 2018. “Where the rubber hits the road" A review of decentralisation in Myanmar and the roads sector". International Growth Centre (IGC) and Renaissance Institute (RI)

Winter, Michael and Mya Nandar Thin. 2016. “The Provision of Public Goods and Services in Urban Areas of Myanmar - Planning and Budgeting by Development Affairs Organisations and Department”. Yangon: The Asia Foundation and Renaissance Institute.

World Bank, 2007. “Intergovernmental Fiscal Transfers: Principles and Practice” edited by R. Boadway and A. Shah

World Bank. 2015a. "Making the hole greater than the sum of the parts: A review of fiscal decentralization in Vietnam".

World Bank. 2015b. “Myanmar Public Expenditure Review: Realigning the Union Budget to Myanmar’s Development Priorities”.

World Bank. 2017. “Myanmar Public Expenditure Review: Fiscal Space for Economic Growth”

https://www.imf.org/external/pubs/ft/wp/2016/wp1614.pdf

https://www.imf.org/external/pubs/ft/wp/2016/wp1614.pdf

https://pefa.org/sites/default/files/assements/comments/MM-Mar12-PFMPR-Public.pdf


34

Appendix

Table A1: Schedule Five and budgeted sub-national taxes

Schedule Five of the Constitution – taxes collected by Region or States 1. *a/ Land revenue

2. *a/ Excise revenue

3. *a/ Water tax and embankment tax based on dams and reservoirs managed by the Region or State and tax on use of electricity generated by such facilities managed by the Region or State

4. *b/ Toll fees from using roads and bridges managed by the Region or State 5. *c/ - Royalty collected on fresh water fisheries

*c/ - Royalty collected on marine fisheries within the permitted range of territorial water

6. Taxes collected on vehicles on road transport and vessels on inland waterway transport, in accord with law, in a Region or a State 7. Proceeds, rent fees and other profits from those properties owned by a Region or a State

8. Fees, taxes and other revenues collected on services enterprises by a Region or a State 9. Fines imposed by judicial courts in a Region or a State including Region Taya Hluttaw or State Taya Hluttaw and taxes collected on

service provision and other revenues

10. Interests from disbursed by a Region or State 11. Profits returned from investment of a Region or State

12. *d/ Taxes collected on extraction of the following items from the forests in a Region or a State: - Taxes collected on all other woods except teak and other restricted hard woods;

- Taxes collected on firewood, charcoal, rattan, bamboo, bird nests, cutch, thanetkha, turpentine, eaglewood and honey-based products

13. *e/ Registration fees 14. Taxes on entrainments

15. *f/ Salt tax 16. Revenue received from the Union Fund Account

17. *e/ Contributions by development affairs organizations in a Region or State concerned

18. Unclaimed cash and property

19. Treasure trove

* Items identified as taxes on state/region budgets

a/ relates to land, excise and embankment taxes collected by General Administration Department (GAD) b/ could relate to wheel tax, which in this instance is a road fee collected by the municipal offices. Collection of fees for the use of non-

municipal highways are collected by a different agency. Interviews with government officials suggest that some of the tolls on state/region highway roads are collected as non-tax fees by the Road Transport Administrative Department, which does not feature on state/region

budgets.

c/ relates to fisheries tax budgeted at state/region level d/ relates to tax on the extraction of forest products budgeted at state/region level

e/ relates to wheel tax, collected by the municipal offices upon registration of road vehicles f/ could relate to mineral tax, collected by GAD and budgeted at state/region level

e/ relates to property tax, collected by the municipal offices


35

Table A2: Budget categories and administrative units

Category Administrative units Administrative sub-units

(departments)

Sub-national government offices

Region and state government bodies

State/Region Cabinet

State/Region Hluttaw (Parliament)

Courts

Legal Offices (Attorney Offices)

Accounting Offices (Auditor Offices)

Ministries


General Administration Department

Department of Special Investigation

Fire Services Department

Ministry of Agriculture, Livestock and

Irrigation

Department of Agriculture

Department of Fisheries

Livestock Breeding and Veterinary Department

Department of Bees

Cooperative Department

Small-scale Industries Department

Ministry of Natural Resources and Environmental Conservation

Environmental Conservation Department

Department of Forestry

Department of Mines

Ministry of Electricity and Energy Electricity Supply Enterprise *

Ministry of Health and Sports Department of Sports and Physical Education

Ministry of Planning and Finance

Planning Department

Central Statistical Organization

Budget Department

Cargo Handling (Management) Office

Ministry of Construction Department of Urban and Housing Development

Department of Highways

Municipal offices

Development Affairs Offices Development Affairs Offices

Yangon City Development Committee

Mandalay City Development Committee Naypyitaw City Development Committee

Notes: Categorization of some departments may differ from one state/region to another and across years. The list above is the

most recent and common representation. * State-owned Economic Enterprise

About the Renaissance Institute

The Renaissance Institute (RI) is a policy institute focused on assisting the reform of Myanmar’s economy. RI

provides analytical support, assists government capacity building and facilitates communication between

Myanmar’s government and other relevant stakeholders focused on revitalizing Myanmar’s economy. In particular,

the Renaissance Institute is supporting key policy priorities of the current government: fiscal decentralization, and

public financial management reform.

No. 51-A6, Thayarwaddy Street

Sayar San Ward, Bahan Township

Yangon, Myanmar

Tel | Fax: + 95 1 555 879

www.rimyanmar.org

Myanmar Sub-national Budget Brief · Ildrim Valley is an economist on the Overseas Development...

Documents

Transcript of Myanmar Sub-national Budget Brief · Ildrim Valley is an economist on the Overseas Development...